Artificial Intelligence

Some sample usecases for customized scenarios of Azue OpenAI for conversational AI :

• Azure OpenAI Virtual Assistant 
• AI Powered Call Center Intelligence Accelerator 
• Enterprise Chat with Azure Search  

Azure AI Bot Service

 Is to design and build enterprise-grade conversational AI bots.

 Amazon Lex

is to build bots with Conversational AI. 

Powered by the same technology as Alexa, Amazon Lex provides you with the tools to tackle challenging deep learning problems, such as speech recognition and language understanding, through an easy-to-use fully managed service. Amazon Lex integrates with AWS Lambda which you can use to easily trigger functions for execution of your back-end business logic for data retrieval and updates. Once built, your bot can be deployed directly to chat platforms, mobile clients, and IoT devices. You can also use the reports provided to track metrics for your bot. Amazon Lex provides a scalable, secure, easy to use, end-to-end solution to build, publish and monitor your bots.

• Natural conversations
• High quality speech recognition and natural language understanding
• Context management
• 8 kHz telephony audio support
• Multi-turn dialog
• Builder productivity
• Visual Conversation Builder
• Powerful Lifecycle Management Capabilities
• One-click deployment to multiple platforms
• Streaming conversations
• AWS service integrations
• Integration with Amazon Kendra
• Integration with Amazon Polly
• Integration with AWS Lambda
• Contact center integrations
• Amazon Connect
• Genesys Cloud CX
• Amazon Chime SDK
• AWS Contact Center Intelligence (CCI) 

 Amazon Bedrock

Designed to build and scale generative AI applications with foundation models. 

Choice of models :

• Amazon Titan
• Jurassic
• Claude
• Llama 2
• Stable Diffusion
• Command 

Amazon Bedrock can be used to build assistants that understand user requests, automatically break down tasks, engage in dialogue to collect information and take actions to fulfill the request.

Language Services

Natural Language Understanding (NLU) is used to extract information from documents, emails, chats and customer communication. Based on the model used the language can be detected, customer sentiment and opinions, text classification. All these information can be used to orchestrate business processes such as prioritize emails for routing, automatic closure of claims, service requests.

The language AI can do following tasks :

• Analyze sentiment and mine opinions
• detect language
• classify text and extract information
• Personally Identifiable Information (PII) detection
• Document summarization
• Conversation summarization

Azure AI Language Studio

Azure Example of sentiment analysis using Azure AI Language Studio using Azure AI Language service is analysing customer complaint about visiting telco shop in the shopping mall.

 {
    "documents": [
        {
            "id": "id__13312",
            "sentiment": "mixed",
            "confidenceScores": {
                "positive": 0.17,
                "neutral": 0.27,
                "negative": 0.56
            },
            "sentences": [
                {
                    "sentiment": "positive",
                    "confidenceScores": {
                        "positive": 0.65,
                        "neutral": 0.34,
                        "negative": 0
                    },
                    "offset": 0,
                    "length": 72,
                    "text": "Dobrý deň, Chcem reklamovať Vašu službu vo Vašej predajni v OC Central. "
                },
                {
                    "sentiment": "negative",
                    "confidenceScores": {
                        "positive": 0.03,
                        "neutral": 0.35,
                        "negative": 0.62
                    },
                    "offset": 72,
                    "length": 75,
                    "text": "Chcel som si kúpiť paušál, no v obchode nebol nikto ochotný sa mi venovať. "
                },
                {
                    "sentiment": "negative",
                    "confidenceScores": {
                        "positive": 0,
                        "neutral": 0.38,
                        "negative": 0.62
                    },
                    "offset": 147,
                    "length": 102,
                    "text": "Keď som prišiel po radu, povedali mi, že vaše systémy sú mimo prevádzky a mal by som prísť ďalší deň. "
                },
                {
                    "sentiment": "negative",
                    "confidenceScores": {
                        "positive": 0,
                        "neutral": 0,
                        "negative": 1
                    },
                    "offset": 249,
                    "length": 22,
                    "text": "som velmi nespokojny. "
                },
                {
                    "sentiment": "neutral",
                    "confidenceScores": {
                        "positive": 0.16,
                        "neutral": 0.84,
                        "negative": 0
                    },
                    "offset": 271,
                    "length": 23,
                    "text": "s pozdravom Jozef Novák"
                }
            ],
            "warnings": []
        }
    ],
    "errors": [],
    "modelVersion": "2022-11-01"
}
 

AWS Amazon Comprehend

AWS Example of  Personally Identifiable Information (PII) analysis mode in AWS Comprehend service

API call

{
    "Text": "Good day,\nI want to complain about your service at your store in OC Central. I wanted to buy a flat rate, but there was no one in the store willing to attend to me. When I came for advice I was told your systems were down and I should come another day. I am very dissatisfied.\nSincerely,\nJozef Novák",
    "LanguageCode": "en"
} 

Result

 { "Entities": [ { "Score": 0.9995710253715515, "Type": "ADDRESS", "BeginOffset": 65, "EndOffset": 75 }, { "Score": 0.9999931454658508, "Type": "NAME", "BeginOffset": 288, "EndOffset": 299 } ] 
} 

 Oracle Cloud Infrastructure AI Service – Language

OCI Example of complex analysis using language detection, text classification, named entity recognition, key phrase extraction, sentiment analysis.